An online platform for spatial and iterative modelling with Bayesian Networks

Abstract

Bayesian Networks (BNs) are commonly used to model socio-ecological systems, as their graphical structure supports participatory modelling, they can integrate quantitative data and qualitative knowledge, and account for uncertainty. Although the spatial and temporal dimensions are important in socio-ecological systems, there is a lack of openly available and easy-to-use tools to run BNs with spatial data over time. Here, we present gBay (gbay.ethz.ch), an online platform where users can link their BNs to spatial data, run the network iteratively to incorporate dynamics and feedbacks, and add geo-processing calculations to account for spatial interactions. We demonstrate the use of this tool on the examples of a modelling a regulating ecosystem service, where we account for neighbourhood effects, and land-use decisions, where we include regional-level boundary conditions. The gBay platform supports users with its graphical interface and instructive wiki page, and provides a step towards more accessible and flexible socio-ecological modelling.